Hey cats. I spent the past 2 days writing a new design/control app for the ez-b and decided to make a new robot as a test platform. so yesterday i noticed i had a few of these bulldozer kits lying around, but i find them annoying due to the DC motor gearbox. A few years ago i started using modified servos instead of dc motors on the bulldozer kit. I get the odd email asking how i did that...
Soooooooooooo, i coupled the two together. A video showing how to build a servo driven tamiya bulldozer and the ez-builder
Any idea of the relative efficiency of continuous servos? IOW, if you want something to go a long ways slowly, are you better off with these servos or sending low voltage to a motor that can go faster?
thanx guys. i use modified servos for everything. higher control and more torque than dc motor gearsets. not sure if you’ve seen videos of my other robots, but i wouldn’t want them to move too quickly anyways.
i don’t think speed was ever a concern with using servos. i beleive programming for servos has always been a nuisance. my project board supports 16 servos, so that’s not a problem.
if you send a small ammount of current to a dc motor, then you end up with very low torque and almost zero precision.
i’m adding a few more bits to ez-builder tonight. i have a few more ideas
Varying the drive voltage to a motor, although intuitively sensible, actually doesn’t work out so well. The speed and particularly torque characteristics don’t scale so well as you adjust the voltage.
Assuming the motor is a brushed DC (and not a synchronous, brushless, etc) it’s much more favourable to leave the voltage supply as high as you can, and use PWM to vary the amount of power delivered to the motor. With the right PWM frequency the speed of the motor scales quite cleanly relative to PWM duty cycle, and the torque remains high even when the speed is low.
It looks like this another situation where I may have already told you more than I know…
When I vary PWM to an ESC that is powering a brushed motor (on my RC truck chassis, for example), I thought it actually varied the voltage that the motor receives.
the result of pwm appears to have a lower voltage, but that’s not true. if you were to hook up an oscilliscope to a pwm output, you would see pulses of the voltage. for example, a PWM of +5 would still result in many quick +5 pulses.
physical hardware that has noticeable resistance, for example a brushed motor or an LED would respond to PWM… as it turns electricity into heat energy very quickly. that’s due to the inefficiency of its design.
CMOS for example, would not recognize PWM as a voltage drop. neither does USB or sensitive logic circuits.
You can affectly “somewhat” maintain torque with PWM and high voltage to a brushed motor. where if you lowered the voltage, you would suffer torque loss.
PWM can be quite magical… but its a lot of work to do it right… when a servo has all of the control embedded in a tiny package. now, that’s not to say you should hook a servo up to a race car unless you want to lose lol
I understand about PWM. PWM is input to the ESC and it is steady 5v. The ESC then delivers power to the motor and depending on which of projects I am working on, the voltage available is either 7.2v, 22v or 48v. I thought what the ESC did was vary the voltage. Further complicating the question is that the big motors are brushless. Only my little 7.2v toy has brushes.
Do the larger brushless motors have 3 power leads? In that case they won’t use PWM anyway, but instead they’ll be running as 3-phase synchronous motors, with the motor speed controlled by the rate at which the phases oscillate. This is quite similar to how many industrial motors and servo motors are controlled. Interesting stuff to read up on if you have the time =)
Yes 3 phase. The PWM I generate goes to the ESCs that control those motors. It’s pretty silly when you think about it. The ESCs are controllers - they have some chip that does logic. But they are programmed to deal with signals from a receiver that is getting analog over RF so they decode that to drive the motors. I have to generate that analog signal even though I have the raw information about how fast I want to drive the motor and in what direction. But I have to do this dance of implication and inference over an analog line instead of just sending digital information. There are digital controllers, but the hardware guys don’t want to use them. They like being able to plug in a RC controller and demonstrate that they have the unit working properly.
thanks for posting.
unless you want to lose lol